RE: draft-allocchio-gstn-05.txt concern

["Steven Legg" <steven.legg@adacel.com.au>]

All,

Claudio Allocchio wrote:
> > That is, I believe this specification is not compatible with
> > RFC2252-defined telephone number syntaxes.  (RFC2252 is on the
> > Standards Track.)
> 
> "Telephone numbers", not dial strings, is it?

Yes, a telephone number. The telephone number syntax in LDAP/X.500
is not intended to be a dial string.

> In the Object identifier
> which is inserted into directories you put telephone numbers, written
> accordingly to the human readable standard E.123, if I'm 
> correct. Thus you
> specify a different object than a dial string, which also contain, for
> example, pauses, or special dial elements like DTMF "A", "B", etc...

My read is that the draft in question does not, and should not, apply
to the LDAP telephone number syntax. For starters, the DTMF '*' and '#'
characters are disallowed characters, and the existing matching defined
for the syntax treats space and hyphen as insignificant, but would regard
'.' as significant.

> 
> Looking, however, from a more pragmatic view, E.123 syntax 
> differs from a
> subset of dial string syntax just because it allows "white space" as
> separator, while dial string syntax, for compatibility with existing
> internet protocol specification and to avoid unduly complex quoting
> mechanisns, uses "." or "-"... not so far away even if 
> they're different
> objects.
> 
> I don't really see the reasons for such a strong position: we 
> created the
> specification just to avoid tha anybody invents his own syntax for a
> commonly used object encoded in text such as a dial string 
> (... and in the
> IETF the "not invente here" often apply till the "not invented in this
> WG", or worse "not invented by me"  (-: ). If a future revision of
> directory specification will simply allow an additional 
> "optional" syntax
> (which in my opinion is not strictily necessary to do, ad 
> you're anyhow
> talking a different object, not a dial string), then you will have the
> problem you mention about implementation compatibility. If you simple
> state "in order to draw a dial string from thei E.123 syntax telephone
> number you need to convert spaces into '.'", than there is 
> not problem.

Yes, the only relevance the draft has to the LDAP syntax is that a
dial sequence can be derived from the syntax by a straightforward
method, which is not normative to the LDAP definition and therefore
does even need to be stated. It would be harmful to interoperability
to impose the dial sequence format on the LDAP syntax.

Regards,
Steven

> 
> Hope this helps to clarify a bit!
> 
> 'night !
> 
> 
> --------------------------------------------------------------
> ----------------
> Claudio Allocchio             G   A   R   R          
> Claudio.Allocchio@garr.it
>                         Project Technical Officer
> tel: +39 040 3758523      Italian Academic and       
> G=Claudio; S=Allocchio;
> fax: +39 040 3758565        Research Network         P=garr; 
> A=garr; C=it;
> 
>            PGP Key: http://www.cert.garr.it/PGP/keys.php3#ca
>